buildbetter-mcp-research

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The required workflow ingests outsider-authored free text only indirectly through the user’s research question/phrases that drive BuildBetter MCP search/list tools (e.g., search-calls, search-signals, search-documents) against organization data, so the agent can read attacker-supplied strings without first selecting a specific pre-existing item.